Background: Rivaroxaban has gained popularity as an anticoagulant (AC) for stroke prevention in nonvalvular atrial fibrillation (afib) and venous thromboembolism (VTE). Although adverse bleeding events are associated with all AC, lack of point-of-care testing to measure the effect of rivaroxaban in emergent situations has contributed to perceived increased risk among physicians.
Methods: This study aims to describe a single-center experience with trauma and emergency general surgery (EGS) patients taking rivaroxaban and evaluate outcomes compared with patients taking warfarin using a propensity score analysis. Trauma and EGS patients taking rivaroxaban or warfarin for afib/VTE over a 2-year period were eligible for inclusion and matched for injury/illness severity in a 1:2 ratio using propensity score matching. In a single quaternary referral center, 192 warfarin patients were matched to 96 rivaroxaban patients. Groups were well matched with no significant difference in age/sex, admission systolic blood pressure/heart rate, admission hemoglobin, injury severity score (trauma patients), or need for ICU admission. Conditional logistic regression determined association of AC type with bleeding complications, adjusting for age/sex, AC indication, coagulation laboratory values, antiplatelet medications or other AC, comorbidities, renal impairment, and operative intervention. Primary outcome was bleeding complications, defined as hemorrhage during admission or as a presenting problem. Secondary outcomes included invasive interventions, AC reversal, VTE complications, and mortality.
Results: There was no difference between rivaroxaban and warfarin for bleeding complications (37% vs. 39%, p = 0.49), VTE complications (4.2% vs. 5.7%, p = 0.44), or mortality (4.2% vs. 5.8%, p = 0.63). Fewer rivaroxaban patients underwent surgical or interventional radiology procedures during admission (32% vs. 43%, p = 0.01), but there was no difference in procedures specifically for bleeding (10% vs. 12% p = 0.68). Rivaroxaban patients less frequently underwent AC reversal (34% vs. 46%, p = 0.01) or received multiple reversal agents (20% vs. 29%, p = 0.02). Regression analysis confirmed AC type was not associated with bleeding complications (rivaroxaban vs. warfarin relative risk 1.02; 95% CI 0.85-1.22, p = 0.85).
Conclusion: Reversal of rivaroxaban was less common and required fewer agents, whereas bleeding complications and hemostatic interventions do not seem to be different between these AC types.
Level of evidence: Therapeutic study, level II.